Does Meta even need frontier research? A two-tweet debate on RSI and open source

NathanpmYoung · x · 2026-09-18

A short AI-circle exchange over Meta: former DeepMind researcher Suchenzang argued Meta doesn't need to solve Millennium Prize problems for its stock to go 'up and to the right.' NathanpmYoung pushed back: if that's the bar, couldn't Meta just ship the latest open-source models? Is the actual technology edge just recursive self-improvement (RSI)?

The brief exchange captures ongoing debate about whether Meta's superintelligence lab needs original breakthroughs or can ride open source to stay competitive.

Related event: Meta Doesn't Need Millennium Problems to Keep Rising, Ex-Researcher Says(2 posts)→

Original post →

More from Companies & People

Companies & People channel →